SwarmAI 3D PathFinding System - Should I sell asset?

Looks great, I bet your game is finished before Star Citizen.

 

Yes it will be, there are advantages to being a solo developer, and that is I know every line of code by heart, which makes it easier to find and fix issues especially with large complex systems.

I also found an artist on the asset store that has every ship i could possibly need for the game already made, exactly what i was looking for an incredibly professionally made. It will cost me a few thousand dollars for all the ships but hey, that's a steal considering how much work he put into them.

My game will also be fun and a true innovation in the space sim genre.

It will NOT be an MMO. It will however eventually be multiplayer assuming the single player version does well. This game will be a real-time 4x strategy game. A Conquest game. Not a game about going and doing stupid missions and fetching like a dog. A game where you lead an empire, you exist in a living persistent universe where your choices REALLY matter and can cause your entire empire to fall apart (full rebellion simulation). A game where death has real meaning and consequence, and most of all a game where Skill matters and a game with innovative "Strong" artificial intelligence that is capable of tactical analysis and adaptation, not just flying in circles.

Might not be as pretty as star citizen but I can guarantee you that it will be a A LOT MORE FUN, well "Actually Fun" because star citizen is amazingly boring since 3.0 release. Such a disappointment. I spent 10 minutes landing on a planet to fetch a box and another 30 minutes traveling somewhere to drop it off and im like this is the most boring game in the world. Sure is pretty though. On a side note, it took me 4 hours to do one mission because it was so buggy, I really do not know how you can mess up such a simple "mission system", as simple as picking a box up lol!! Not like there is some real complicated mechanics behind that.

I hope at some point people here acknowledge that I'm not just blowing steam and I am capable of building the game, I spent years learning and mastering the engine to get to where I am now. People like to get you down around here, I get it though.

Check out my latest vid: I'm not really great at "communication" so I'm not very good at commentating on videos. I am good at programming however and that is what matters right?!

Oh and here are the assets I will be using, it's a large collection that has pretty much everything you could ask for, jump gates, pirate fleets, asteroid stations! Using these assets will seriously bring my development time down because i will not be needing to wait on modelers.

I can build the concept and if it takes off, eventually i will have some custom models made, for that I'll need some serious money, as you know.

https://www.assetstore.unity3d.com/en/?stay#!/search/page=1/sortby=popularity/query=publisher:1611

 

I would buy one ship first before committing. I got this ship from another good modeler: https://www.assetstore.unity3d.com/en/?stay#!/content/60206
It was disappointing. The ship looks great from afar but the textures get very blurry up close. I think any model on the asset store will suffer from that. You will need a very good LOD system, or you will have to use Vertex Texturing.

 

Just found out you can actually cut a model up and have LOD's for different parts of the model. Then again, no ship on the asset store is gonna be set up like that. They are all more for top down games.

 

I am not to concerned about textures, if it is an issue ill use standard or some uber metallic shaders. You can view his models in sketchfab. But yah, I'll buy some and check them out, but you can tell the guy is a professional artist by his work.

 

Yah there are assets that can do it as well but, you have to keep in mind lod systems use resources as well. I would only use it if the number of verts and draw calls is an issue. Building a game is a delicate dance Balancing on a tight-rope while dancing at the same time.

 

I know. I've tried making a

1. VR RPG ala Skyrim:

Abandoned due to my partner putting in no work for months and me realizing it would take decades to finish
2. VR Survival game set in the Jungle:

Abandonded due to not reaching desired plant density and desired framerate
3. VR shooter:

Abandoned due to reaching the limits of playmaker (I'm an artist not a programmer)
4. Online Hovertank shooter with tactical elements:

This is the first time I actually started coding. This was also abandoned due to me being the only member on the team to put in work for months

And now I'm at a simple top down space shooter:

I know i know, those kinds of games are a bunch a dozen, I just want to get a game done. And believe me there will pop up all kinds of unexpected things even with simple stuff like this.

 

Hey that is some nice work! I made a lot of prototypes when i was learning, not nearly as pretty as yours (I'm just a programmer). (Architecture and AI systems design is what I find most fun) I don't use play maker, I write all of my code from scratch, my own controllers and framework. My only limitation is my imagination and ability to properly organize the code and systems (which is why architecture design is SO important when dealing with large complex systems).

You should see what I am building, people will call me insane. But, hey I probably am a little..

Attached is a rough draft of the architecture I am building for Galactic Armada. I have entered the design phase now. Since I have proof of concept with the collision avoidance path finding system that I designed, which was probably the most difficult component of the game to build.

The AI is more time consuming than "difficult". It's really fun to design and build though. Luckily, I can spend 15+ hours a day every day on it for the next 3 years

The chart is not done yet, I will have it completed tomorrow. Before I begin building it though I have about 5 other flow charts to create regarding the Galaxy System, Ship Spawning System & Pooling System, Economy System, Board Game Battle Simulator (Background process) and the
Empire Leader AI system which is a separate AI from the one in the flow chart attached. The one in the flow chart runs ships.

Hey I could use an artist! You should check out what I'm building. Really after i get the full set of flow charts done you will be able to see what the game "IS" which is a 4x realtime strategy CONQUEST game inside of a "physical 3D Galaxy" with epic space action combat, roaming AI (FLEETS) - Pirates, Marauders, Space Creatures and ENTITIES!

You can see some of the prototypes I made when I was learning on my old channel.


Realistic storm with procedural lightning, volumetric clouds.

An earlier version of the storm system but on the ocean, waves grow large ect.

I wanted to fly over a rainbow to the moon. Somewhere over the rainbow.

Is that a physics simulation inside a moving boat?

Oh there a whole bunch more and.. a bunch i never even posted to youtube. My aspirations in game development are in space though, not on the ground, at least for the next few years.

Going to build a great space sim. Something new for once and not an MMO. I don't think "playing fetch" is all that fun. Thinking is fun and so I'm building a game for people who like to think and also like epic space action combat and strategy.

 

Hey with your jungle, your issue there is batching, without question. You need to combine all of those plant models into solids, unless you are using "wind", in that case it's a problem. If you can sacrifice having wind like speedtrees use, you can combine all of those mesh into a single mesh and reduce those draw calls. Optimally you don't want saved by batching to be over a few hundred or at least under 1K. Also, if you are using occlusion with so many draws it can actually have a negative effect on performance in my experience.

 

Any Questions

Next Chart is galaxy system. I know my chart isn't exactly "pretty". I don't have a week to spend on each chart. My next chart will look better, using a new chart editor.

 

I have not done much work in 2d but i would imagine it is FAR, FAR more simple than a 3D system. My system for example has about 4,000 lines of code in 6 different classes just for the collision avoidance system. I doubt it would take nearly that much for a 2d system. Though, my system is incredibly sophisticated and allows for all kinds of behavioral steering modifications and routines as well as algorithms for specific scenario detection. (Like when ships get stuck and need to turn around) Or detect multiple ships and geometry (moving) blocking it's path.

 

For the Jungle vegetation I used https://forum.unity.com/threads/critias-tree-system-asset-store.437520/
Indeed I was using wind and a day/night cycle. It was just too much. Will definitely revisit that project at some point.

The space shooter is actually 3D, just set in the XZ plane with a fixed camera view. My AI is laughably simple. Here it is:
The main class, a simple go to point:
https://hastebin.com/ulicapezah.cs
then I made some simple behaviors deriving from that class:
Wander: https://hastebin.com/vazemomoqe.cs
Flee: https://hastebin.com/majowepetu.cs
Attack: https://hastebin.com/aducepimez.cs
As you can probably see from that code I'm a bloody beginner, been coding for about 3 months now. I'm sure this code will lead to wacky and stupid behavior by the AI but to be honest I don't think that good AI is that important, on the contrary, if the AI is too good then the player will get it's ass handed to him too often and fun will go down.
But that is of course a different story in a 3D RTS. If X3 had better AI and fleet management and ran better I'd still be playing it. That's basically the game you are trying to recreate.

 

AI is not everything though. Have you thought about story, setting, history, characters, plots, economy? Sound effects, Music, voice syncing, UI, VFX? Character Animation, IK, rigging, lip syncing? Camera smoothing and clipping, first person, 3rd person views. Funding, Advertisement, what kind of company, LLC or C-Corp? Contracts, IP, hiring?

 

Nah my game will be entirely different than X, the way you give "orders" will be different, the fleet system much more comprehensive and designed. When X was designed, it was not designed around "fleets" it just kind of became part of the game in the sequels. My game won't have a comprehensive trading or economy system as it will be based upon the "Risk" model, where planets generate resources over X time period as well as "spawn ships of certain classes". Players will be able to purchase ships as well as will be given ships through the spawning system.

There will be few resources, Galactic Currency, Ammunition, Energy Cells & Fuel. These resources will be generated by planets and supply ships will spawn and travel to fleets and deliver supplies.

You are not the first to make that statement about AI, I disagree. The game I am building will be a dangerous place where death is a real possibility, and dying to often will cause your empire to fall apart because your ships will rebel against you. If my game does not kill you at least a few times during the cycle of a game, even if you are a skillful pilot, then I have failed in my design. I'm not saying im going to force it to be impossible, I am saying I'm going to make it incredibly difficult to make it through the game without ever being in a situation where you can be killed if you make a mistake.

My game will also have very simple interfaces for creating and assigning ships to fleets as well as giving orders to fleet admirals. X's Gui's were way to convoluted with too much of a learning curve. The game was hard to get into, it was fun though once you did, if you and when you made it past the learning curve.

My game will have much more sophisticated AI as well as a "living galaxy" in constant flux and much more simplified interfaces which will make it a lot easier to get into and play, within minutes.

Also there will be entities in GA that will be able to permanently alter the galaxy by destroying planets and entire solar systems.

Anyway this will also be detailed in the charts. I'm working on the procedural galaxy generation system and the economy system charts now which detail both. I don't mind sharing them when they are done.

 

Yes I have.

1. I have owned several companies
2. I am a full stack web developer (LAMP) with 5 years of experience in social media marketing and SEM lead-gen marketing and 15 years of experience in web development as a specialist in ad-network technology among other things.. Magento, Facebook API integration (FQL) (I have connections).
You forgot to mention the website, but I will say that I have that covered. I will be starting on it within the month and it will be modern and professional, it will also use a custom framework + CMS system that I designed in PHP.
3. GA won't have a story, it is a sandbox game with emergent gameplay. There will however be "random events" within the galaxy as well as end game crisis (similiar to stellaris).
4. There will not be any character animations in GA, at least not initially.
5. Sound effects are easy, there are numerous FX libraries where you license just about every sound FX imaginable.
6. Music also easy
7. there will be very few voice overs in GA, at least initially.
8. there will be no first person views, the game will be experienced from the cockpit of space ships, similiar to X in that regard.
9. Did you watch my last few videos on the channel???
10. I have funding for everything that I need to build the game as well as much time as I need to build it without worrying about any economic concerns.
11. Also have an advertising budget.
12. Company is LLC
13. This is a solo gig, until it isn't. I would be willing to take on a partner and share the load if they are capable of performing development tasks, have a passion for space sims and believe in my project.
14. As far as the camera system, I have already designed it and you can see it functioning in the videos on the channel, of course it could use some improvement with smoothing in the 3rd person view with the external ship chase camera, just have not got around to it yet. Also keep in mind the videos are recorded at 30FPS and the game is running well over 150+ FPS, so what you see in the video isn't exactly what the render looks like. The only way to smooth it out for the recording is to increase the recording FPS which massively increases the video size and takes me forever to upload or to enable VSync to bring fps down to 60.

 

Any plans for base building, expanding and improving planets? With such a simplified economy I'm not sure how living that galaxy would appear. I don't think you are thinking about the different types of players and how some would cheese the system. I'm the type of "cowardly Admiral." I'd hide behind the frontlines and just command my fleets. I'd play it as a pure RTS which makes me wonder if your game would be better as a straight up RTS.

 

Well damn, from your videos I was under the impression you were pretty young. That's a hell of a resume. I have no coding or computer related experience whatsoever. All self taught in my spare time. Very little free time to work on stuff too. As much as I'd love to work on such a grand project I just realistically can't and I'm not sure I'd be a real asset. But I would like to contribute at least one larger model of some kind. I never worked on a really large model before so it'd be a new challenge.

 

No because the empire leader being present in battles will give bonuses to all ships present. Technically you will be able to stay behind your lines and give orders to fleet admirals and manage your empire from the galaxy map. However, the longer you loiter without involving yourself the more likelihood you will lose battles which will effect morale, which can cause rebellion over time. There will also be roaming fleets, pirates and marauders, plenty of activity where the player will not have enough resources to deal with them all without being involved in combat.

You have to be realistic about what one person can build. I can't succumb to feature creep like star citizen. I have a core set of features that I must have that revolve around the types of fleets and entities the galaxy.

With that said the economy system will be simplified as well as the procedural galaxy, which you will see in my mechanics charts in the coming week. These systems of course are being designed to be "Expandable". But initially, I cannot have a complex economy system, or an incredibly complicated procedural galaxy generations system.

I can however have basic systems that are expandable. Building blocks for more complex systems.

GA won't be an incredibly sophisticated RTSG like stellaris. It can't be. because of resource limitations. It will be a simplified version.. more focus will placed on fleet combat, ship combat within the 3d element of the game initially where the RTS systems will be expanded later after players "alpha test" the concept with a fully playable prototype.

You can hit my up on skype if you like: elitephpcoder

 

Edit: I find the idea of the Ship Devourer really interesting. You could make it so it swallows the ship whole, internally it get's boarded and re-crewed and sent out the back again as your ship.

 

No planet upgrades or base building, though those could be expansions. You have to understand that my goal is not to build all of these "ideas" initially.

My goal is to build the core experience within the galaxy initially which includes the simple economy system, the fleet(s) and ship combat. The galaxy map empire presentation (displays fleet positions and activity). The various AI systems.

You do not want to do what star citizen has done which is try to build everything and then find out that you completely wasted your time on alot of systems and end up removing them.

My strategy is the opposite which is the design the core elements which MUST be in the game. The game will actually be playable with these core elements. Then have people play it and give feed back, and only after that add additional systems and improvements which will add to the experience.

These additional systems could be the very things you mention such as planet upgrading, defense platforms even station weapon systems upgrading like in stellaris. These things are not absolutely necessary for a playable game for people to experience the concept behind the game.

 

Actually, they will be creatures like blowfish with large mouths and they will Physically chase your ships and when they get within 30 meters they will suck the ship into the mouth using Physics (gravitational pull).Once swallowed they will destroy the ship, they will be deadly. They will stream "feces" (Particle stream). After they devour.

It will be hilarious. My game will not be without a sense of humor. When you get one behind you, you have 2 choices, turn and fight or run. Keep in mind, they won't care what direction your ship is pointing in, so you better prey you kill it before it swallows you whole.

I'm telling you, my vision for this game is like nothing you have ever seen before. I know this for a fact as a super-space-sim-fan boy that has played them all.

That's not where the humor ends, if you die and alot of your ships have a low morale already, you will respawn to your own empire trying to murder you. Death will have a massive negative negative morale modifier empire-wide.

Also death will effect your clones if you die too much you will start "glitching out" for random periods of time including "Fast blackouts, fuzzy vision, steering bumps" and more.

The rebellion simulation system will be a MAJOR part of this game and a lot of time and effort will go into it. It will also be detailed in a chart, along with all of the mechanics. Which I will have all of it down in the next 2 weeks covering all mechanics for the entire game.

 

The more you describe everything, the more I think it's doable. Honestly had my doubts it was too grand of a project. Any plans for modding? I know that's a pain to implement, but could really help fill out the game for you.

 

I am using the mega pack ship assets from this artist:
https://assetstore.unity.com/packages/3d/vehicles/space/progenitor-fleet-mega-pack-i-75098

I'll be purchasing over 1,000$ worth of his ships and other assets in fact. The ships are perfect for my needs.

Not needing to wait on modelers to create and design models will significantly, if not insanely reduce development time for the game, which means getting a playable game will mostly depend on how fast I can write, deploy and test the software.

this is where the website will come into play. The website will have a forum as well as an alpha testers section where forum members will be granted access to the current alpha build as well as individual test builds which contain isolated mechanics for testing. Free of charge, no donation required.

The only requirement being testers give feed back as well as upload a TXT file which will contain test data relative to system specs and performance recordings as well error logs.

Part of the development for this game will be community driven where i listen to feedback, that only comes after I have my core elements for the game in place and functional though so that players can at least in part experience and understand the basic concepts behind the game.

Yes this game is very doable. I already have a proof of concept with the galaxy system and the game can be built with the unity engine using floating origin systems and asynchronous scene loading. The galaxy map will be a scaled down version of the procedural generated galaxy that will rotated to fit on 2d board.

 

oh modding, not modeling..

No, I don't think so. I wish but again.. you have to be realistic about what one person can build and that is feature creep. If the game takes off and people like the concept, something like that could be a possibility in a sequel or an expansion where there is more than just me doing all of the programming work.

All of my programming systems are modular in design so integrating a modding system would be a possibility. But given the size, scope and complexity of some of these systems .. a modding system would be alot of work.

I have thought about that though, with such a system people would be able to turn my galaxy into a star wars galaxy or a star trek galaxy if they had the models, even babylon 5.

A SHIP MODDING SYSTEM..
would not be much work at all. Allowing people to replace ship models.. would be extremely easy actually.

But a modding system for everything else.. alot of work and additional testing time.

 

Galaxy Map / Procedural Galaxy Concept Design Diagram for Galactic Armada.

Economy system overview will follow since it goes hand-in-hand with the galaxy, since planets and "areas" will generate resources, it is just too much to fit onto one chart.

 

Anyway, I'm done updating this forum thread.

If you would like to follow development and checkout my charts as i release them, follow this thread:
https://forum.unity.com/threads/galactic-armada-fleet-commander-overview-mechanics-videos.516871/

After the design phase I will be building the website where these systems and mechanics will be better explained in a way a non-software-engineer can understand.

I expect the design phase to last for several weeks, hopefully i can get all charts done within 3 weeks. The entire game will exist "on paper" within these charts. As well as alot of software design flow charts for each system.

 

Still mulling ever putting up your swarmAI on the Asset store or even sharing it on GIthub? It looks to be a very impressive system! Kudos on a job well done.

 

Nah.. I got a job building a game, then it turned into 2 games.. now im so busy i can't even think about that lol.

I may end up building a new version of my AI system for a game I am working on though.